Create User Personas

"First-job earners aged 21 to 28 who are paid monthly" is a precise target actor, but it is hard to design for a category. In a debate about a new screen, someone says "users will love this", someone else says "users will hate it", and both are imagining a different person. A persona gives the whole team the same person to picture — one grounded in research.

The rule

A persona is research in human form; every detail must change a decision.

A persona is a fictional but evidence-based portrait of a group of target actors. It has a name, a short story and the facts that matter, so the team can ask "would Riya do this?" instead of arguing from personal taste.

Where it comes from matters most. A useful persona is built from research with real actors — interviews, surveys, usage data — and summarises patterns you actually found. A persona invented in a meeting just gives the team's assumptions a face.

For behavior design, the most useful details are the ones from the target actor lesson:

• Goal for the target outcome, in their own words.
• Context: how they are paid, where and when they use their phone, what their month looks like.
• Prior experience with the target action — including past attempts that failed, and why.
• Motivations and barriers around the action.
• Relationship to the product: how they found it, what they expect from it.

Keep the set small — one to three personas, each tied to the target actor — and revisit them as research grows.

The shortcut

Run the "so what?" filter on every detail: "So what would we design differently because of this?" If the answer is nothing, cut the detail. If you cannot trace it to research, check it or drop it.

The trap is a persona full of vivid, invented trivia. It feels real, but it steers the team towards details nobody has checked and away from the ones that should shape the design.

Drills

Answer fast — trust the rule.

Remember

• A persona is research in human form; every detail must change a decision.
• Shortcut: the "so what?" filter — cut details that change nothing or that no research supports.
• Trap: vivid, invented trivia that feels real and steers the team wrong.